PM-QPSK相干光通信系统恒模算法研究

发布时间：2018-04-23 05:06

本文选题：PM-QPSK + 相干光通信　；参考：《天津理工大学》2015年硕士论文

【摘要】：偏振复用正交相移键控(PM-QPSK)相干光通信系统不仅具有相干光接收技术的高灵敏度、长距离传输等特点,而且结合了偏振复用技术和多进制调制格式(正交相移键控),进一步提升了通信系统容量,因此,PM-QPSK相干光通信系统受到越来越多的关注。除此之外,在长距离传输中信道的非理想特性(色度色散,偏振模色散)可以在接收端由信号处理技术进行补偿。本文研究了PM-QPSK相干光通信系统中用于信道均衡的恒模算法。主要研究内容包括:(1)提出了一种时域变步长恒模算法,该算法具有迭代步长自适应变化的特点。已有的恒模均衡算法中通常采用固定步长,对步长的选择通常需要在均衡速度和均衡误差之间寻找平衡。而在本文提出的均衡方案中,均衡的步长会随着算法的运行自适应的更新。本文对影响步长自适应更新的算法关键参数进行了分析和选取,仿真结果表明,在稳态误差相近的条件下,本文算法收敛所需的迭代次数相比于传统恒模算法约减少600次,占总迭代次数的18%。在误码率为10-3时,与传统的恒模算法相比光信噪比(OSNR)代价下降了约1.7d B。(2)提出了一种低复杂度高精度恒模均衡算法,该算法结合了时域均衡高精度和频域均衡低复杂度的特点。本文算法首先利用频域恒模算法对输入信号进行粗均衡。随后以误差函数为基准,对每个信号点频域粗均衡的效果进行判断,并对其中性能较差的信号点进行重新均衡。为了尽可能地减小计算复杂度,重新均衡采用了固定系数时域恒模均衡方案。仿真结果表明,所提算法相对于时域恒模算法,其计算复杂度减小了6%;在误码率为10-3时,与时域和频域的恒模算法相比OSNR代价分别下降了约2d B和2.2d B。
[Abstract]:Polarization-multiplexed orthogonal phase-shift keying (PM-QPSK) coherent optical communication system not only has the characteristics of high sensitivity of coherent light reception technology, but also long distance transmission. The PM-QPSK coherent optical communication system has been paid more and more attention because of the combination of polarization multiplexing technology and multi-ary modulation format (orthogonal phase-shift keying) to further enhance the capacity of the communication system. In addition, the non-ideal characteristics of the channel (chromatic dispersion, polarization mode dispersion) in long distance transmission can be compensated by signal processing technology at the receiving end. In this paper, a constant modulus algorithm for channel equalization in PM-QPSK coherent optical communication systems is studied. This paper presents a time-domain variable step size constant modulus algorithm, which is characterized by the adaptive variation of iterative step size. The existing constant modulus equalization algorithms usually use the fixed step size, and the choice of the step size usually needs to find a balance between the equalization speed and the equalization error. In the proposed equalization scheme, the step size of the equalization is updated adaptively with the running of the algorithm. In this paper, the key parameters of the algorithm which affect the adaptive updating of step size are analyzed and selected. The simulation results show that the number of iterations required for convergence of the proposed algorithm is about 600 times less than that of the traditional constant modulus algorithm under the condition of similar steady-state error. 18 of the total number of iterations. When the BER is 10 ~ (-3), compared with the conventional OSNR algorithm, the cost of OSNR is reduced by about 1.7 d B. ~ (2). A low complexity and high precision constant modulus equalization algorithm is proposed, which combines the characteristics of high precision equalization in time domain and low complexity in frequency domain. In this paper, the input signal is coarsely equalized by frequency domain constant modulus algorithm. Then based on the error function, the effect of coarse equalization in frequency domain of each signal is judged, and the signal points with poor performance are rebalanced. In order to reduce computational complexity as much as possible, a constant modulus equalization scheme with fixed coefficients in time domain is adopted. Simulation results show that the computational complexity of the proposed algorithm is reduced by 6 compared with the time-domain constant modulus algorithm, and the OSNR cost is reduced by about 2dB and 2.2 dB, respectively, compared with the time-domain and frequency-domain constant modulus algorithms when the BER is 10-3.
【学位授予单位】：天津理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN929.1

【相似文献】